Understanding Financial Vulnerability in Partially Dollarized Economies

The reduction of macroeconomic vulnerability in emerging markets is now at the core of the research agenda. Liability dollarization plays a vital role in the understanding of vulnerability and its implications (from a general equilibrium perspective) have been addressed in the literature via the inclusion of a “financial accelerator†mechanism. In particular, its formalization is based on Bernanke’s, et al. (1998) optimal contract, which predicts a negative relation between an external finance premium and firm’s net worth. We can identify two channels by which the financial accelerator can be triggered. The first, emphasized in Bernanke, et al. (1998) and Gertler, et al. (2001), operates via shocks on asset prices which, in turn, affect the realized return on capital and net worth. The second channel, privileged in Céspedes, et al. (2000a y 2000b), depends on unanticipated movements in firm’s debt burden which directly affect their net worth. Not surprisingly, liability dollarization plays an important role in the activation of this second channel since the unexpected component of a real depreciation can greatly magnify the debt burden of firms if their debt is denominated in dollars. Based on this, Céspedes, et al. (2000a y 2000b) present a first approximation to a definition of vulnerability. In particular, an economy is classified as vulnerable if a real exchange rate depreciation implies an increase in the risk premium faced by firms. This result is neatly summarized in the log linear version of the risk premium equation and depends, crucially, on firms indebtness level. Their model, however, assumes complete depreciation and, thus, lacks the asset price channel explained above. Gertler, et al. (2001) recognize this issue and present some simulations using dollar denominated debt and an active asset price mechanism. Despite these significant contributions to the understanding of the consequences of liability dollarization for output fluctuations, we believe some important extensions are now in order: (i) if we want to address the implications of the degree of dollarization, we need a general equilibrium model that admits firm’s debt to be denominated in both local and foreign currency (the two models just described assume full liability dollarization); (ii) central bank’s response to exchange rate innovations (given a degree of dollarization) must be assessed from a welfare point of view; and (iii) given a dollarization level and central bank’s response to shocks, a new, encompassing, definition of vulnerability must be provided in order to adequately address the way in which it can be mitigated